Another missed opportunity on UK tech policy
December 2019
Unlocking the value of London’s public sector data
Back in 2013, as a think tank newbie, I took my first forays into the weird and wonderful world of digital government.
One observation struck me almost at once: while changing the nature of the technology that public sector organisations use (towards common standards, interoperability, government-as-a-platform and so on) is likely to be the work of a couple of decades, every organisation has an abundance of data, and we can get cracking on unlocking its value right now.
To achieve that, I’ve long argued that it’s not enough for individual organisations to make more rigorous use just of their own data.
That’s for two reasons.
First, while data is no cure-all (see Thea Snow’s recent article for an excellent explanation on why), many of the smarter ways of working that public sector organisations wish to make possible depend on data collaboration.
For example, if we want neighbouring councils to be able to spot and have productive conversations about where they might share teams, assets and resources on issues of common concern, it helps to have data on how those issues manifest themselves on both sides of their boundaries.
If we want multiple public sector organisations to intelligently coordinate their work (an imperative in areas such as social care), they need data on what each other is doing.
And if we want to enable prediction and prevention, or prediction and earlier intervention, we need the datasets that can collectively point to cases of highest risk.
Second, and more fundamentally, in a big, complex city like London, challenges, opportunities and people’s lives don’t neatly confine themselves to within one borough’s boundaries. So neither must our data.
Several well-known factors make data collaboration hard.
There are technical barriers. Some legacy systems don’t play nicely with others, making it hard to get the data out. The lack of a common technical platform on which to share data can also be a hindrance. That’s why Paul Neville at Waltham Forest and Trevor Dorling at Greenwich are ensuring LOTI is fully engaged in the GLA’s discovery phase on the future of the London Datastore, which may well become the default platform for data sharing.
More troubling, some IT suppliers charge local authorities thousands of pounds to extract or create an API to their own data. In my view, no large supplier who does that can seriously claim to be in the business of helping local government. I wrote that it was bad practice in 2013. Today, it should be completely unacceptable. At LOTI, we’re using City Tools to provide a better evidence base to inform which systems provide the best value.
There are data barriers. Records can be collected according to different standards and conventions. A lack of unique identifiers like Unique Property Reference Numbers (UPRNs) makes linking records challenging.
There are cultural barriers. If organisations aren’t in the habit of demanding data to inform their actions and decisions, why would they go to the effort of building the supply?
There are skills barriers. We need good data scientists and analysts, yes. But effective data collaboration depends on many more roles, from technical specialists to information governance professionals. It’s also helpful if all staff have at least a base level of understanding of how better data might enhance their activities.
Finally, there are legal barriers, both real and perceived. The real ones are there for good reason. There are some things you shouldn’t be able to do with data, and rightly so. The perceived barriers operate on two fronts and are more problematic. First, some data initiatives that could provide real value to citizens never happen because public sector staff assume they can’t be done. Second, when each organisation has a different approach to ensuring that their data sharing initiatives are conducted legally, ethically and securely it can create months of delay or even halt worthwhile projects from ever getting started.
LOTI boroughs recognise that addressing this is a vital part of laying the groundwork for future success with data.
It’s for that reason that we’ve been working with CIOs, Data Protection Officers and Information Governance leads to work on creating a common approach to information governance that could be followed by boroughs. In a project led by Ed Garcez and his colleagues at Camden, we’ve started sketching out a seven-step IG process, describing what needs to happen and who needs to be involved at each step. We’re trialling the use of the Information Sharing Gateway, a platform that digitises and standardises the process of creating information sharing agreements (ISAs). And we’re looking carefully at the potential of collaborating with the Greater Manchester Combined Authority on an initiative with CC2i to create a digital Data Privacy Impact Assessment.
There’s also more work to be done on open data.
The open data world began with a focus on bringing transparency to government. Yet arguably its greatest impact and future potential are in the realms of promoting innovation.
The poster child of open data innovation is Transport for London. TfL’s ingenious decision to release its Unified API, offering machine-readable, real-time data from across its infrastructure and services, rather than developing apps itself, has led to the creation of 700+ apps, products and services by external developers.
While a significant part of that success is linked to the fact that transport data lends itself to real-time data and has some fairly obvious consumer-facing use cases (journey planning apps etc.), it holds at least one important lesson for local government.
If we want to prompt real innovation, we need to release open datasets at scale.
Put plainly, developers need a larger potential customer base than the residents of just one local authority area to have a viable business model. In the absence of scale, local authorities are likely to find their data used by hobbyists, at best.
The need for scale also comes from citizens. For local authorities that take user needs seriously — and that covers all that have signed the Local Digital Declaration — we should acknowledge that the user need is not always as a resident of a borough but as a Londoner. While some services rightly have a distinct local character, others do not, and the average Londoner does not want to have to use 33 different parking apps, and so on.
In short, we need a more strategic conversation about what datasets, currently held by local government, really need to be released consistently as open data at a London level in order to provide the products, services and value that both residents and local authorities desire.
There are two ways to have that conversation.
Data is a means, not an end. Ideally, we’d therefore start by thinking about what we’d like to be able to do that we cannot currently achieve, and then explore what might help enable those things. Personally, I love GovLab’s 100 questions initiative. “What 100 questions, if answered, would have the greatest positive impact?” We could try that framing in London.
Alternatively (or perhaps, additionally), I think it’s legitimate to ask whether there are some datasets that are so fundamental to the operation of a modern, thriving city that they should be regarded as our vital “data infrastructure”, to use the ODI’s phrase.
I recently posted this question on Twitter — and twitter wasn’t short of ideas.
Similar questions and considerations are likely to inform our thinking on the Internet of Things (IoT).
As the EU’s Sharing Cities programme has demonstrated, if implemented in a thoughtful manner, IoT-enabled devices can provide significant value to cities and their residents, making buildings more energy-efficient, monitoring traffic flow and air pollution, and so forth.
Yet before London boroughs start implementing this kind of smart street infrastructure in earnest, I think we need to be clear that there’s a future we want to avoid. One in which each borough procures its own, closed IoT networks, collecting data in inconsistent standards that can’t be shared, and with citizens left in the dark about what data is being collected and why.
In short, we want to design out the problems and data silos we’ve experienced with legacy technology. We’re currently scoping out a project with Nathan Pierce from the GLA and Ben Goward (Westminster & RBKC) to achieve that. More details to follow soon.
Recognising both the opportunities and challenges sketched out in this article, LOTI boroughs are now working on a Joint Statement of Intent on Responsible Data Sharing to set out their ambitions, principles and approach to data collaboration. There are many factors to consider, with the need to use data in an ethical and transparent manner that engenders citizens’ trust and confidence being top of the list. We’ll share a draft shortly.
These are just a few of the things that LOTI boroughs are working on to help unlock the value of London’s public sector data. It’s not easy and won’t be achieved overnight. But getting this right might well be one of the most important things we can do together.
LOTI is hosted by London Councils.
Originally published here.
